National Institute for Space Research
Biblioteca Digital da Memória Científica do INPENot a member yet
6175 research outputs found
Sort by
Comparação entre HOG+SVM e Haar-like em cascata para a detecção de campos de futebol em imagens aéreas e orbitais
No full textAutomatic object recognition in digital images is not an simple task due to diverse variations present within this process, consequently, different general purpose techniques have been proposed. In this paper, an approach combining HOG and SVM and other utilizing Haar-like cascade for automatic soccer field detection in airborne and satellite imagery is analyzed.Pages: 8154-816
Análise de tendências em séries temporais de dados meteorológicos e dados de sensoriamento remoto orbital em áreas de vegetação campestre natural do bioma Pampa, localizadas na Depressão Central do RS
No full textThis work aims at checking for trends in time series of meteorological and remote sensing data in areas of natural grassland, located in Depressão Central,RS. The analyzed period was from 04/01/1998 until 12/21/2011. This study used ten-day period averages of minimum, mean and maximum air temperature data, ten-day period of accumulated rainfall and ten-day period of accumulated evapotranspiration data provided by an automatic weather station, located in Eldorado do Sul-RS (30\ub006'20''S, 51\ub041'10''W). Blue, red, near infrared and mid-infrared bands and Normalized Difference Vegetation Index (NDVI) of Vegetation sensor was used. Six areas of native grassland were identified to be analyzed by the Vegetation sensor. Plots were constructed to show the dispersion and linear trend line using Sigmaplot 10.0 software. Correlation analyzes with a ten-day period lag between the meteorological data and the Vegetation sensor was performed. The accumulated rainfall was the only element of climate that tended to increase. The other meteorological data showed stability during the analyzed period. For the Vegetation sensor data, the most sensitive bands to changes in the amount of water in the canopy, as the mid-infrared and red bands, showed negative trend. This result is consistent with the increased cumulative rainfall. In correlation analyzes with a ten-day period lag, the band near infrared and NDVI showed higher correlation coefficient with the meteorological data.Pages: 3003-300
Uso de dados MODIS e AIRS para obtenção de parâmetros de correção atmosférica
No full textRemote sensing by spectral imaging of the Earth's surface can be widely used to characterize the surface objects composition, but only when the atmospheric influence is corrected and the data are transformed to surface reflectance units. If atmospheric correction is performed using the MODTRAN radiative transfer code, users have to supply input parameters in order to characterize the atmospheric conditions of the imaging flight time. Since these parameters are difficult to obtain, especially in tropical regions, we propose a methodology to estimate them from MODIS atmospheric products. We used aerosol optical thickness (τ), Ångström exponent (α) and H2O total column data in order to retrieve initial visibility (V), retrieve aerosol scale height (za) and select the fittest MODTRAN atmospheric model to be used in the atmospheric correction. CO2 mixing ratio complementary data were retrieved from AIRS and AMSU sounders onboard the Aqua satellite. As results, the atmospheric correction improved reflectance values and widened the NDVI values range of the corrected images. The correction produced good attenuation in shorter wavelength visible spectral bands reflectance values and achieved augmentation in longer wavelength infrared spectral bands reflectance values. These results point out that the parameterization has enhanced the atmospheric correction procedure, generating more reliable surface reflectance information from satellite images.Pages: 8019-802
Mudanças no uso e cobertura da terra no Maciço da Tijuca/RJ: as implicações nas áreas de preservação permanente entre os anos de 1972 e 1996
No full textThe Maciço da Tijuca is a geomorphological system in Rio de Janeiro, where are located the Serra de Três Rios, Serra da Carioca, a Pedra da Gávea and the National Park of Tijuca. The area that was covered by a dense Forest, stated to be deforested to supply civil construction, the olarias, in addition the use for the coal industry and the coffe farming. After that, the forest was replaced by buildings due to urban expansion that, as in most Brazilian cities, happens disorderly, disrespecting the boundaries of Permanent Preservation Areas (PPA). This occupation brings problems caused by the pressure performed in PPA, since they play important environmental functions, like regulate of the water flow, sediment retention, ecotourism, soil and biodiversity conservation. The use of remote sensing and geoprocessing to detect and characterize the land use and land cover changes is important for monitoring the landscape dynamic, mainly in large areas. Thus, this work approaches the Permanent Preservation Areas in rivers, nascent and slopes, through the land use and land cover change detection in Maciço da Tijuca, for years 1972 and 1996, using the fill cell of TerraView 4.0.0. The PPAs limits were stated using the brasilian Forest Code (1965) and the 303/2002 CONAMA Resolution. As results, were produced maps that indicate the areas where the uses and cover are in disagreement with the law.Pages: 7140-714
Identificação de feições morfológicas de praias através de Imagens de Satélite Quickbird
No full textThe coast morphology is the result of a set of variables among which are the hydrodynamic processes, geological and climatological. Morphological variations that occur with beach profile due to seasonal changes in wave energy, are the product of an intense and dynamic displacement of sediment, and the knowledge of morphology praial by monitoring a stretch of beach and its adjacent background checks maintaining a tendency erosive or progradational. To conduct the study, we used sensor Quickbird images of 2005. It was observed using image embayment, tidal flows, cusps (rhythmic beaches), rip currents, spikes forming features of erosion and sediment accumulation. It was concluded that this method allows simple and rapid identification of morphological features and morphodynamic praiais, and can also be used as a tool for identifying risk to bathers.Pages: 5422-542
Aplicação da lógica fuzzy para identificação das condições de moradias urbanas de Ponta Grossa-PR
No full textFuzzy logic is implemented in the various sciences, especially for its efficiency in modeling problems in which the variables have subjective parameters. This paper makes use of fuzzy logic to the IBGE 2010 census data in a GIS environment to represent their housing conditions and access to water, garbage, sewage, electricity and the number of households without a restroom. The selected variables and spatialized in census tracts have the living conditions of the population of the city of Ponta Grossa - PR. through indexes that show changes from 0 (highest level of absence) to 1 (lowest level of absence) of housing conditions. The best values are present in the Neighborhood Center and the conditions are more unfavorable in the peripheral sectors of the city. The end result of applying fuzzy indices showed satisfactory results, reflecting the housing conditions pontagrossense in relation to the lack of census tracts in relation to other sectors.Pages: 786-79
Aplicação de espectroscopia de reflectância VNIR-SWIR na quantificação de atributos do solo através de estatística multivariada
No full textDiffuse reflectance spectroscopy have been used to analysis soil attributes improving the quality and the quickness result. This technique reduces the environment degradation because dont use chemicals and dont eliminate residuals. The electromagnetic radiation in different wavelengths can be detected with spectral sensor and give us the reflectance curve. With this information we can describe the curve and your features, also compare soil attributes with respective reflectance information. For this, chemometrics help us doing the statistics analysis, data manipulation and figure out the results. This work analyzes the soil attributes and reflectance information (VNIR-SWIR) with Partial Least-Square Regression aiming quantifies the soil attributes with remote sensing. Thus, soil data was collected in central region of Mato Grosso do Sul, Brazil and analyzed in routine laboratory and remote sensor laboratory. FieldSpec Pro (Analytical Spectral Devices, Boulder, Colo.) was used to measure the spectral reflectance in visual, near-infrared and shortwave-infrared bands. To run the statistic method, PLSR, was used the software ParLeS. This software model data to predict soil attributes with the reflectance curve. Model of clay, silt and sand was acceptable with respective values: R²=0,9325; R²=0,833; R²=0,9175; RPD=3,87; RPD=2,46; RPD=3,50. For organic matter, calcium, magnesium and sum of base the results was semi-quantitative (OM: R²= 0,6906; RPD=1,49; Ca: R²=0,7988; RPD=1,86; Mg: R²=0,7035; RPD=1,83; SB: R²=0,7746; RPD=1,85).Pages: 498-50
Uso de imágenes SAR Cosmo-Skymed para la determinación de humedad del suelo
No full textSoil moisture-SM influences the redistribution of radiant energy, runoff generation, infiltration, and therefore, recharge. Knowing its spatial distribution is strategic for agricultural activities. In situ measurements of SM are time consuming, providing only point information. Synthetic aperture radar SAR images bring the possibility of estimating SM over extensive areas, with adequate spatial and temporal coverage. In this work, preliminary maps of SM were obtained by applying an Artificial Network Algorithm (ANN) in order to invert the values of the Advance Integral Equation Model (AIEM). Five X band SAR images from the satellite COSMO-SKYMED were processed and calibrated. Soil moisture was measured in situ by means of a TDR probe, samples were taken from the field for gravimetric water content determination. Days for field campaigns were performed as close as possible to the image days. The comparison between SM and backscattering values shows dispersion, in particular under very humid conditions. However, the multitemporal analysis showed increasing backscaterring values as SM increases, as expected. The application of the ANN algorithm showed consistency between field data and model simulated values. This work was developed in the State of Santa Fe, Argentina.Pages: 5729-573
Paleorias fluviais associadas ao rio Madeira, sudoeste da Amazônia brasileira
No full textThis paper discusses the use of remote sensing to improve the morphological description of open vegetation patches and ria lakes in two distinct areas on the Brazilian Amazon. JERS-1 images, SRTM-DEM and multispectral Landsat TM-5 images were used to extract water and vegetation features, which were discretized in groups of pixels using an object-oriented approach. Several attributes regarding geometry and spatial distribution of lakes, ria lakes and open vegetation patches were calculated. Six attributes were identified by the InfoGain algorithm as the most pertinent to individualize each group. The major shape similarity between ria lakes and open vegetation patches suggests a common origin to these features. The distribution of lakes in the Solimões River floodplain also indicates a major shift in this river course toward the South. By comparison with the ria lakes of the Solimões River, it is also suggested that the Madeira River shifted its course to the East and during this process, the left margin tributaries were progressively abandoned. These ancient fluvial morphologies have provided a different geological substrate in which the open vegetation has growth. Previous studies within the Brazilian Amazon have also related open vegetation patches with depositional sites that have been active during the late Quaternary.Pages: 3728-373
Mapeamento preliminar de conflitos de uso da terra na bacia do Rio Pardo - MG
No full textThe growing global demand for natural resources and food has exerted pressure on the environment and has reflected negatively in the geographical space occupation process, with increased of deforestation rates and natural areas degradation. In this context, legal mechanisms were created with the objective of assigning sustainability to the process of human activities expansion. One such mechanism includes the Permanent Preservation Areas. However, as the anthropic land occupation precedes this mechanism, the scenarios of illegality and spatial conflicts are already established. For these situations, mapping of APP's and land use/cover can identify and quantify conflicts, generating information for adopting actions of adequacy and territorial reorganization. In this sense, the objective of this study was to evaluate the conflicts of land use in the watershed of the Rio Pardo - MG. The methodology consisted of applying the following steps: i) generation of land use/cover mapping;, ii) generation of Permanent Preservation Areas mapping; iii) overlay operation between the land use/cover and Permanent Preservation Areas maps; iv) identification, mapping and quantification of conflicts in land use for the study area. It was observed, preliminarily, that the Permanent Preservation Areas occupy approximately 27% of the study area, and 59% of this total may be in conflict with any human activity, especially Forestry.Pages: 6246-625